Older adults demonstrate deficits, relative to young adults, in explicit tests of memory, such as recognition tests, as well as tests of working memory. In addition, epidemiologic and neuroimaging studies suggest that there is differential susceptibility to age-related memory changes that is related to variables such as education , IQ, and engagement in leisure activities. We have hypothesized that there are two complementary facets to reserve against the effects of aging: Cognitive reserve describes the normal individual differences in the capacity to perform tasks. This differential capacity, might result in some people being less susceptible to the effects of aging than others. Compensation is the use of alternate brain networks not normally used by younger individuals as a response to the effects of aging. The proposed research is aimed at exploring the neural mechanisms that underlies age-related memory deficits and the differential reserve against these changes. We have three key questions: 1) What are the neural systems that underlie variability in task performance in young adults? 2) Do healthy elders use these same systems, or do they use alternate compensatory systems? 3) How does of the use of these systems relate to factors that have been associated with reserve, such as IQ and education. We propose to delineate these neural system with five fMRI cognitive activation studies. The specific aims are to: 1. In young and elderly subjects, identify brain networks whose expression varies as a function of task load on two tasks, delayed match to sample task (Sternberg task, working memory), and continuous non-verbal recognition (recognition memory). 2. Explore network changes as performance is challenged by manipulations that affect the difficulty of specific aspects of task processing. 3. Compare expression of these load-sensitive networks in young and elderly subjects, to determine which neural networks underlying task performance are similar, and which change as a function of aging. 4. Evaluate how response to task load varies as a function of variables known to mediate cognitive reserve, including IQ and education, in order to identify neural networks associated with cognitive reserve and compensation.